THE RITUAL opens an AI-assisted social-deduction prototype for testing
By GameSlash
The creator is testing the core loop before dedicated art and audio. We tried voting and Reveal with bots, without claiming an eight-human playtest.
Phattanapong Panpetch has opened THE RITUAL for testing, describing it as an AI-assisted social-deduction project. Eight souls form a circle and seek the hidden outsiders through discussion, voting and powers used at night. This is a prototype testing its core systems—not an announcement that development is complete.
For AI game creators, the interesting choice is to test a playable system before completing dedicated visuals and audio. The developer says those assets are not yet made specifically for the game and that a full session with eight human players has not been tested. The immediate questions are whether the rules are clear and whether the social loop makes people want another round.

Vote by day, use asymmetric information at night
The in-game rules assign five Drifters, two Fallen Drifters and one Morph. Daytime discussion leads into a vote: the player with the most votes is cast out, while a tie spares everyone. At night, roles use their powers before the next dawn reveals the consequences.
Drifter powers include Fortune, which checks whether a target is Human or Ghost, and Shield, which can protect against a night attack. Fallen powers include Reveal, exposing a Drifter's skill, and Mute, silencing a target. Players therefore see different information. The Morph can disguise its result during the first two nights and has a special win condition when three players remain.

The bot test verifies actions, not an eight-person social session
GameSlash created an isolated room with one tester and seven built-in spirit bots, without microphone use or invitations to other people. The first fast-timer round assigned Drifter–Fortune and ended in defeat on day three. A second, normal-timer round assigned Fallen Drifter. We selected Wraith, confirmed the ballot and saw the vote recorded.
On the first night of that normal-timer round, we selected Reveal on Wraith and confirmed it. At dawn on day two, the private result read 'Wraith holds Fortune.' This verifies a target–action–result sequence in actual play rather than merely showing a menu.

This test covered phase transitions, voting and some ability behaviour. It did not verify a smooth eight-human voice session or establish game balance. The creator also warns that the test bots are less capable than real players. The current presentation emphasises player circles, status text and an event log rather than finished visual production.
A prototype with a clear testing question
The useful distinction for creators is between a functioning system and an enjoyable human session. Selecting a target and receiving a result helps test the first question; conversation flow and deception still need human playtests. That is an observation from this prototype, not a claim that AI immediately delivers a finished game.
See the creator's post dated 30 October 2026 and try THE RITUAL. It remains in development and may contain bugs.